Individuals with PKU lack the enzyme phenylalanine hydroxylase that is needed to metabolize the essential amino acid phenylalanine (phe). When eating a normal diet they show an elevated level of phe in blood that is toxic to the brain. In order to prevent brain damage and cognitive impairment, individuals with PKU must follow a lifelong, low-phe diet that is restricted in natural foods and requires ingestion of a phe-free amino acid (AA) formula. Most adolescents and adults with PKU find the AA formula unpalatable and go off the diet resulting in elevated blood phe levels and neuropsychological deterioration. The risk and potential health care costs are especially high for pregnant women with PKU whose infants show congenital anomalies when exposed to elevated maternal blood phe levels, i.e., maternal PKU. Glycomacropeptide (GMP), an intact protein produced during cheese making, is uniquely suited to a low-phe diet because it is the only known dietary protein that contains minimal phe. Foods and beverages made with GMP are a palatable alternative to AA formula. The long-term goal is to assess the safety, efficacy and acceptability of GMP for the nutritional management of PKU. The specific aim is to conduct a randomized, two-stage, 11-week (wk), crossover trial comparing the GMP diet with the AA diet in 30 subjects with PKU =12 years of age treated since birth with a low-phe AA diet. The sites are: University of Wisconsin- Madison, Waisman Center (primary) and Harvard University, Children's Hospital Boston. Subjects will be recruited and randomized to begin the first 3-wk of the study with either a low phe diet in which the majority of dietary protein is provided by GMP or AA medical foods and then, after a 3-wk washout with intake of their usual diet, begin the second diet for 3-wk. Dietary education will be provided in a 1-wk base period preceding initiation of each diet. The primary endpoint is change in the plasma phe concentration of PKU subjects fed the GMP diet for 3-wk compared with change in the plasma phe concentration when fed the AA diet for 3-wk (paired t-test, pairing on subject). Secondary outcome variables include compliance assessed by food records, acceptability assessed by questionnaires, and cognitive function assessed by neuropsychological tests. The investigators expect that compared to the usual AA diet, the GMP diet will reduce plasma phe levels and improve cognitive function due to improved dietary compliance and phe utilization. To enhance understanding of the etiology of low bone mass, a common long-term complication in those with PKU, exploratory data will be obtained to assess markers of bone turnover in a subset of 20 PKU subjects, 12-20 years of age.